Skip to content

Amazon

Principal Applied Scientist, Agentic Outbound Communications Marketplace

Amazon is hiring a Principal Applied Scientist, Agentic Outbound Communications Marketplace in Seattle, United States. It pays $199k-$269k a year and Level rates it ; you can apply on Amazon.

Level's read on this job

  • Builds AI ●●●● is how Level rates this applied scientist role: the job is building AI systems.
  • Pay on this posting starts 19% higher than the typical applied scientist posting on Level.
  • The posting mentions AI agents among its tools and skills.
  • Amazon has 20,000+ other open roles on Level.
  • Posted in the last few days.

AI in this role

ai-agents
Outbound is re-imagining the communications and marketing landscape with advanced generative and agentic AI, changing how millions of customers discover products and engage with brands across Amazon.com and beyond. We're a passionate group of innovators building responsible, intelligent AI that balances the needs of brands and shoppers, improves the shopping experience, and strengthens the marketplace.

Amazon reaches shoppers billions of times through email, mobile push, SMS, and other messaging channels. Today, which message a shopper gets, on which channel, and when is decided mostly by marketers and category owners scheduling campaigns. We're replacing that with a marketplace. Agentic marketers and category owners are the demand. Each shopper's limited, per-channel attention (an email slot, a push notification, an SMS) is the supply. An auction decides which message earns each attention unit, and an autonomous research loop keeps improving the system against shopper engagement.

As a Principal Applied Scientist, you'll set the scientific direction for this marketplace and own the models and methods it runs on:

- Market and mechanism design. Define the value-based bid currency (predicted value to the shopper and the business, not cash) and the allocation mechanism across shoppers, channels, and send times. The mechanism needs to reward honest value reporting by agents, respect hard shopper-protection constraints, and keep any one category from crowding out the rest.

- Shopper value, response, and fatigue modeling. Build the models that price an attention unit: open, click, and conversion prediction; unsubscribe and complaint risk; send-time and channel affinity; and the long-term cost each message imposes on a shopper's future attention.

- Simulation science. Lead the science behind a high-scale synthetic environment of marketer agents and shoppers. Calibrate it against production, measure the sim-to-real gap, and use it to find bidding strategies that game the auction before real shoppers see them.

- Autoresearch and autonomous optimization. Following Andrej Karpathy's autoresearch pattern, agents run closed experiment loops: propose a change to scoring, allocation, or send-time policy, test it in simulation and then on a small slice of live traffic, score it against a fixed objective, and keep or revert the change. You'll define the objective (CTR with guardrails), the evaluation methodology, and the statistical rules that let the loop run without fooling itself or eroding shopper trust.

Key job responsibilities
- Set the multi-year science roadmap for the agentic communications marketplace and align science, engineering, product, and marketing leaders around it.
- Design and analyze the allocation mechanism and bid currency, including incentive properties, budget and quota allocation across categories and programs, and behavior when every participant is an LLM agent.
- Build and launch shopper response and fatigue models that work at Amazon scale and in real time across email, push, SMS, and in-app channels.
- Define the marketplace objective. CTR is the primary signal, but the objective has to guard against the obvious failure mode, a system that sends more, clickbait-style messages, by accounting for unsubscribes, complaints, deliverability and sender reputation, downstream conversions, and long-term shopper engagement.
- Lead causal and counterfactual evaluation (holdouts, long-term effect estimation) so offline and simulated results reliably predict online impact.
- Develop the methodology for autonomous experimentation: experiment budgets, multiple-testing and sequential-testing controls, overfitting protection, and safe rollout and rollback criteria.
- Apply reinforcement learning, contextual bandits, and constrained optimization to send-time, channel, and allocation decisions.
- Publish internally and externally, file patents, and represent Amazon in the research community on agentic marketplaces and AI-driven communications.
- Mentor scientists and engineers and raise the science bar across Outbound.

Basic qualifications

- 5+ years of hands-on work in predictive modeling and analysis experience
- PhD in Electrical Engineering, Computer Science, Mathematics, or a related technical field
- Experience working in predictive modeling and analysis
- Experience distilling informal customer requirements into problem definitions, dealing with ambiguity and competing objectives
- Experience programming in Java, C++, Python or related language
- Experience with leading experienced scientists as well as having a record of developing junior members from academia or industry to a career track in a business environment

Preferred qualifications

- 10+ years of relevant work in industry or academia experience
- Knowledge of problem solving, algorithm design and complexity analysis
- Experience creating novel algorithms and advancing the state of the art
- Have peer-reviewed scientific contributions in premier journals and conferences

Amazon is an equal opportunity employer and does not discriminate on the basis of protected veteran status, disability, or other legally protected status.

Our inclusive culture empowers Amazonians to deliver the best results for our customers. If you have a disability and need a workplace accommodation or adjustment during the application and hiring process, including support for the interview or onboarding process, please visit https://amazon.jobs/content/en/how-we-hire/accommodations for more information. If the country/region you’re applying in isn’t listed, please contact your Recruiting Partner.

The base salary range for this position is listed below. Your Amazon package will include sign-on payments and restricted stock units (RSUs). Final compensation will be determined based on factors including experience, qualifications, and location. Amazon also offers comprehensive benefits including health insurance (medical, dental, vision, prescription, Basic Life & AD&D insurance and option for Supplemental life plans, EAP, Mental Health Support, Medical Advice Line, Flexible Spending Accounts, Adoption and Surrogacy Reimbursement coverage), 401(k) matching, paid time off, and parental leave. Learn more about our benefits at https://amazon.jobs/en/benefits.



USA, WA, Seattle - 198,900.00 - 269,000.00 USD annually

How we rate this

Principal Applied Scientist, Agentic Outbound Communications Marketplace at Amazon rates 84 out of 100 for how much of the daily work is AI. That makes it Builds AI (AI Level 4 of 4). The level is about AI in the job, not seniority.

Classification

Builds AI. The job is building AI systems.

  1. ●●●● Builds AI80 to 100
  2. ●●●○ Works on AI60 to 79
  3. ●●○○ Uses AI40 to 59
  4. ●○○○ Little AI0 to 39

Levels come from how often the tools, models and workflows of the role are named in the posting itself. Read the description and count the AI tools, models and workflows it names.

Prepare for this job

Built from this posting and from similar jobs on Level.

Skills and AI tools this role asks for

AI agents

Also common in applied scientist jobs at this level

Pay

This job starts at
$199k
Typical applied scientist job
$167k

This posting starts 19% above the typical applied scientist job. See applied scientist jobs

Questions you could be asked

  1. How do you decide when an AI agent can act on its own versus asking for approval first?
  2. How would you decide a model or AI system is ready to ship?
  3. Tell me about a time a model underperformed in production. How did you find out, and what did you change?

Adapt your resume

  • List these exact terms on your resume: AI agents. An applicant tracking system matches the wording, not the idea.
  • Attach one line of real, concrete experience to at least one of them. A tool named with nothing behind it rarely survives a human read.
  • Lead with what you built, trained or shipped. This role is judged on the AI system itself, not the tools around it.

Get your CV reviewed before you apply

Send your CV and the role you want next. Your review arrives by email usually within a few hours.

Get new applied scientist jobs (Builds AI ●●●●) by email

One email a week, each job rated for how much AI is in the work. No recruiter spam, unsubscribe in one click.

Free.

Similar roles

Research roles that build AI, at other companies.

What kind of AI work fits you?

Answer 12 practical questions in about three minutes. Get a simple profile, the work it points to, and live roles to explore next.

Find my next step

More jobs at Amazon

More applied scientist jobs

Related searches

Same AI level

Jobs by city